Harvard University - Department of Molecular & Cellular Biology

GEOMETRIC CUE FOR PROTEIN LOCALIZATION

by Kumaran S. Ramamurthi and Richard Losick

March 6th, 2009


Kumaran S. Ramamurthi and Richard Losick

Although bacteria are much more architecturally simple than eukaryotic cells, it has been evident for over a decade now that they are very highly organized on a molecular level. For example, bacteria are able to distinguish between different subcellular regions that display no evident chemical uniqueness and routinely sort proteins to these locations in order to establish polarity or build localized supramolecular structures and machines. In the absence of any obvious chemical landmarks at these sites, the mechanisms by which these destinations are identified by bacterial proteins have been mysterious.

We have been studying the assembly of a thick protective protein shell that surrounds bacterial spores. Spores mature as a smaller cell inside a larger cell. The proteins that make up the shell are deposited by the larger cell onto the surface of the smaller, inner cell. Each of these proteins finds its way to this location by a identifying a preceding protein that localized there, but how does the very first protein identify this site? The surface of the inner cell is the only convex, or positively curved, surface inside the larger cell. We have found that the first localizing protein of the shell discriminates between slightly convex and concave surfaces, and preferentially embeds in convex membranes. Altering the morphology of the cells so that convex surfaces are eliminated abrogated the proper localization of this protein. However, when convex surfaces were restored, the protein resumed its localization there. In fact, this protein localized preferentially to convex surfaces even when expressed in other organisms like yeast. The last piece of evidence that finally convinced us came from an experiment that was done in close collaboration with Sigolene Lecuyer, a postdoctoral fellow in Howard Stone's laboratory in Harvard's School of Engineering and Applied Sciences. When incubated with a mixed-size population of large membrane vesicles, in the absence of any other cellular factors, the purified protein preferentially adsorbed onto the smallest, more convex vesicles, suggesting that membrane curvature alone can drive the localization of this protein. We are now trying to understand the molecular mechanisms by which this geometric cue is sensed and are investigating whether sensing membrane curvature is a conserved strategy for protein localization.

Read more in Science

View Richard Losick's Faculty Profile